Probabilistic-Programming-and-Bayesian-Methods-for-Hackers
Probabilistic-Programming-and-Bayesian-Methods-for-Hackers copied to clipboard
chapter 1 - running on mac metal m2 with tfp - run into error InvalidArgumentError: 730 nodes in a cycle
Hi. I'm new to TensorFlow and TensorFlow Probability. I'm running the examples in chapter 1, and when specifying the posterior sampler I get this error when running graph_sample_chain in line 45:
InvalidArgumentError: 730 nodes in a cycle [Op:__inference_graph_sample_chain_5335]
I'm running it on a Mac with M2 Pro, metal and all TF dependencies have already been installed. Any idea of how this could be fixed?
Hi. Another update. I tried running it in Colab with an A100 (to make failure fast) and it hung. At least running it M2 throws the error quickly. Any ideas? Thanks!